The Unsinkable Carrier: Midway Atoll Under Attack, 3–4 June

Burning oil tanks on Sand Island after the Japanese carrier air attack on the morning of 4 June 1942. U.S. Navy photograph, National Archives, public domain, via Wikimedia Commons.

The Battle of Midway · Chapter 3 of 9

The Unsinkable Carrier: Midway Atoll Under Attack, 3–4 June

How Midway's patrol planes, Marines and Army bombers found the enemy, survived Tomonaga's raid and struck back at a terrible price, 3–4 June 1942.

2 May – 4 June 1942Sand Island · Eastern Island · the ocean west and northwest of Midway~25 min read

Midway Atoll was the object of the whole Japanese plan, and for two days it was also the American side’s most exposed piece of the battle. The carriers of Task Forces 16 and 17 could hide in the open ocean northeast of the atoll. Midway could not move. Its two low islands, its airfield and its seaplane base sat exactly where Japanese planners expected them, about 1,100 nautical miles northwest of Pearl Harbor. Everything stationed there in early June 1942 was meant to find the enemy first, to warn the fleet, to absorb the first blow and then to strike back.

This chapter follows the atoll’s garrison from Admiral Chester W. Nimitz’s inspection visit on 2 May through the morning of 4 June. It covers the patrol planes that made first contact on 3 June, the Army bombers and Navy flying boats that made the first attacks on the approaching invasion force, the Japanese carrier raid that struck Sand and Eastern Islands at 0630 on 4 June, and the attacks Midway’s own aircraft launched against the Japanese carriers between about 0705 and 0820. Those attacks cost the garrison heavily and hit nothing. Why they still mattered is one of the questions the chapter tries to answer with the evidence rather than with the legend.

A note on the evidence. Much of what Americans believed about these hours in 1942 and 1943 came from aircrew claims made in good faith under fire. The Office of Naval Intelligence’s Combat Narrative of 1943 repeated many of those claims. Postwar Japanese records and testimony, used by the Army Air Forces historians Wesley Frank Craven and James Lea Cate, by the Marine Corps historians, and later by Jonathan Parshall and Anthony Tully in Shattered Sword (2005), cut the claims down sharply. Where the sources disagree about numbers, this chapter gives the figures side by side and says whose they are.

Times are Midway local time (Zone plus 12), as in the American reports; Japanese records kept Tokyo time, so Japanese accounts give the same events different clock times and sometimes a different date.

Two islands inside a reef

Midway is a coral atoll near the northwestern end of the Hawaiian chain: a roughly circular barrier reef enclosing a shallow lagoon, with two islands of sand and scrub on its southern side. Sand Island, the larger and more westerly, held the seaplane base, the hangars and ramps for the patrol planes, the fuel tanks, the old cable station and most of the living quarters. Eastern Island, smaller and flatter, lay across a channel to the east and carried the airfield, a triangle of runways that served the Marine squadrons and the Army bombers. In 1942 neither island rose more than a few feet above high tide over most of its area.

The atoll had been American since the nineteenth century. A Commercial Pacific Cable Company relay station on Sand Island linked it to Honolulu and Guam from the early 1900s, and in the 1930s Pan American Airways used the lagoon as a stop for its trans-Pacific clipper flying boats. The Navy turned the place into a naval air station during 1940–41, dredging the channel and lagoon and building the Eastern Island field. Japanese destroyers shelled Sand Island on the night of 7 December 1941, the same day as the Pearl Harbor attack, and Japanese submarines surfaced off the atoll several times in the following weeks. By the spring of 1942 the garrison knew it was a target.

The command arrangement was divided in the way of most island bases of the period. Commander (later Captain) Cyril T. Simard, USN, commanded Naval Air Station Midway and was the senior officer on the atoll. Colonel Harold D. Shannon, USMC, commanded the 6th Defense Battalion and was the senior Marine, responsible for the ground and antiaircraft defense. The aviation units, Marine and Army and Navy alike, worked from Simard’s air station; the Army bombers that arrived at the end of May remained Seventh Air Force aircraft.

Nimitz comes to Midway, 2 May

On 2 May 1942 Admiral Nimitz flew to Midway in a PBY-5A and spent the day inspecting every installation on both islands, according to Lt. Col. Robert D. Heinl’s official monograph Marines at Midway (1948). The visit came while the codebreakers at Pearl Harbor were still assembling the picture of a Japanese operation aimed at a target designated “AF”; the identification of AF with Midway was not yet settled. Nimitz’s inspection was a commander’s look at a place he expected might have to hold, and it was followed within weeks by a stream of reinforcements.

What followed was a rapid thickening of the defenses. Heinl records that on 25–26 May a 37-mm antiaircraft battery of eight guns from the 3d Defense Battalion, under Capt. Ronald K. Miller, arrived together with Companies C and D of the 2d Raider Battalion under Capt. Donald H. Hastie and 1st Lt. John Apergis. A 3d Defense Battalion 3-inch antiaircraft group under Maj. Chandler W. Johnson, and a platoon of five light tanks, came in aboard the aircraft transport Kittyhawk. In all, three 3-inch batteries (12 guns), a 37-mm battery (eight guns) and a 20-mm battery (18 guns) were temporarily attached from the 3d Defense Battalion to Shannon’s command.

The 6th Defense Battalion’s own armament included 5-inch seacoast guns in Batteries A and B, 3-inch antiaircraft batteries, 7-inch and 3-inch naval guns emplaced for coast defense, searchlights and many .50-caliber machine guns. Shannon’s Marines laid mines and wire along the beaches, dug in the guns and built shelters for the command posts. Heinl’s monograph does not give a single strength figure for the battalion at the battle, and none is offered here. The ground force was, in any case, preparing for a landing it would never have to fight: the decisive test would come in the air.

The air garrison

Marine aviation on Midway was organized as Marine Aircraft Group 22. Maj. Ira L. Kimes took command of the group on 20 April 1942, replacing Lt. Col. Wallace, and on the same date Maj. Lofton R. Henderson replaced Maj. Chappell as commanding officer of Marine Scout-Bombing Squadron 241. The group’s fighter squadron, VMF-221, was commanded by Maj. Floyd B. Parks. Heinl’s count of Marine aircraft as of 31 May was 21 Brewster F2A-3 Buffalo fighters and seven Grumman F4F-3 Wildcats, together with 19 Douglas SBD-2 Dauntless dive bombers and 17 Vought SB2U-3 Vindicator scout bombers.

The Marine fighters were the weak point and everyone knew it. The F2A-3 Buffalo was outclassed by the Japanese Type 0 carrier fighter in climb, turn and speed, and the handful of F4F-3 Wildcats, though better, were too few to change the balance. The scout-bombers were also uneven. The SBD-2 Dauntless was a sound design, but many of VMSB-241’s pilots were new to it. The SB2U-3 Vindicator was an older fabric-covered aircraft that the Marines regarded as obsolescent.

Army Air Forces bombers arrived in the last days of May. Craven and Cate record that six B-17s of the 26th Bombardment Squadron flew to Midway on 30 May and that on 31 May six more came from the 431st Squadron, two from the 31st and one from the 72d. Lt. Col. Walter C. Sweeney Jr. of the 431st led the B-17 force in action. Two other squadrons on their way to the South Pacific, the 18th Reconnaissance and the 69th Bombardment, each contributed two Martin B-26 Marauders rigged to carry torpedoes. Heinl lists 17 B-17s and four B-26s on Eastern Island as of 31 May.

The Navy contributed the patrol planes and six new Grumman TBF-1 Avengers. The Avengers were a detachment of Torpedo Squadron 8 that had been training on the new type when the rest of the squadron sailed in Hornet with its old TBD Devastators; the detachment was led by Lt. Langdon K. Fieberling. The story of the parent squadron, which lost all fifteen of its TBDs later on the morning of 4 June, is told in Chapter 4 and on the Torpedo Squadron 8 unit page. The PBY Catalinas, flown by crews of several squadrons of Patrol Wing 2 (see the Patrol Wing 2 page), were Midway’s long-range eyes. Their number depends on the count used: Heinl gives 16 PBY-5As on 31 May, while Craven and Cate count 30 PBYs by 3 June.

3 June: first contact

Forewarned by the codebreakers, Midway’s PBYs flew searches out to hundreds of miles along the probable Japanese approach routes every morning from late May. On 3 June they found the enemy. The ONI Combat Narrative records the first surface contact at 0904: a patrol plane reported two Japanese vessels bearing 247 degrees from Midway at 470 miles. The NHHC photographic history identifies the pilot as Ens. James P. O. Lyle and the ships as the minesweepers Tama Maru No. 3 and No. 5, which had sailed from Wake on 31 May.

Within the half hour, another PBY found something much bigger. Ens. Jewell H. “Jack” Reid of Patrol Squadron 44, flying PBY-5A 44-P-4, came on Tanaka’s transport group about 700 miles west of Midway. The ONI narrative renders his report this way:

The text above is the report as given in the ONI Combat Narrative (1943), chapter 4, timed there at 0925; This Day in Aviation places the sighting shortly before 0900. Reid’s crew believed they had found the Japanese main body. In fact they had found the Midway Occupation Force: transports and escorts, not the carriers. The narrative records that by 1100 the same plane had counted 11 ships on course 090 at 19 knots before it was ordered home for lack of fuel. Nimitz, who knew from the codebreakers that the carriers would come from the northwest, warned that the force sighted to the west was not the striking force. The carriers in Task Forces 16 and 17 held their positions northeast of the atoll.

The first American attack of the battle followed that afternoon. According to the ONI narrative, nine B-17s were ordered out at 1100, each carrying bomb-bay fuel tanks and four 600-pound bombs, and took off about 1230 under Lt. Col. Sweeney. They found the transport group about 570 miles out at 1623 and bombed in three flights of three. Craven and Cate give the same nine aircraft, 36 bombs of 600 pounds and a bombing height of 8,000 feet; the ONI narrative has the flights at 8,000, 10,000 and 12,000 feet. The NHHC photographic overview, by contrast, says six B-17s made the attack.

The crews reported hitting warships and transports. The Seventh Air Force claimed five direct hits and several near misses; the 1958 Marine Corps history records claims of “two battleships or heavy cruisers” and two transports hit. Neither claim survived the postwar evidence. Craven and Cate concluded that the claims needed “a radical scaling down,” that one transport probably took a hit that caused a small fire, and that the warships were undamaged; the Marine history says flatly that the bombers hit nothing. It was the pattern of the next two days in miniature: brave, long-range, high-altitude attacks against maneuvering ships, honest claims, and very little damage.

1. 0904 · 3 JunEns. James P. O. Lyle’s PBY reports two Japanese vessels 470 miles west-southwest of Midway; they are minesweepers from Wake.
2. c. 0900–0925 · 3 JunEns. Jewell H. Reid’s PBY reports the “main body” 700 miles west; it is the Occupation Force transport group.
3. 1230 · 3 JunNine B-17s under Lt. Col. Walter C. Sweeney Jr. take off to attack the transports.
4. 1623 · 3 JunThe B-17s bomb the transport group 570 miles out; claims of several hits are later reduced to, at most, one minor hit.
5. 2115 · 3 JunFour torpedo-armed PBY-5As take off from Midway for a night attack.
6. c. 0115–0130 · 4 JunThree PBYs attack the transport force by moonlight; one torpedo hits the oiler Akebono Maru.

The night torpedo attack

The second attack on 3 June was an experiment. Four PBY-5A amphibians were each fitted to carry one Mark XIII aerial torpedo, a capability the flying boats had never used in combat, and at 2115 they took off from Midway to find the transport group by moonlight. The ONI narrative identifies the leader as Lt. William L. Richards, executive officer of Patrol Squadron 44; the NHHC photo caption notes that he flew with VP-24, and that the other pilots came from VP-24 and VP-51.

Cloud banks broke up the formation; two planes lost contact with the leader around midnight and 0100. At about 0115 on 4 June the remaining aircraft found roughly ten ships in two columns with destroyer escort. Richards dropped at about 800 yards on the largest ship he could see. Lt. (jg) Douglas C. Davis dropped at about 200 yards without seeing a result and then strafed. Ens. Gaylord D. Propst believed he saw a hit before a Japanese aircraft attacked him. Ens. Allan Rothenberg failed to find the force and turned back. Propst later ran out of fuel and landed at sea near Lisianski Island; he and his crew were down for 53 hours before rescue.

In 1943 the ONI could say only that “one or two transports or cargo vessels were possibly damaged.” Postwar records showed that one torpedo had struck the oiler Akebono Maru, which nonetheless kept station with the convoy. NHHC’s H-Gram credits the hit to Propst’s aircraft, timed there at 0153, and calls it the only successful American aerial torpedo attack of the entire battle. That distinction is worth holding onto: in the next nine hours, Navy, Marine and Army aircraft would launch dozens of torpedoes and bombs against Japanese ships from Midway and from the carriers, and not one more American aerial torpedo would hit.

Dawn, 4 June: “many planes heading Midway”

Before dawn on 4 June Midway put its search planes into the air on the usual fan to the west and northwest. The ONI narrative records that the B-17s cleared the field about 0415, initially to attack the transports again; they were diverted to the carriers once these were found. At 0430, about 240 miles northwest of the atoll, Vice Adm. Chuichi Nagumo’s four carriers launched the Midway strike: 36 Type 97 carrier attack planes armed as level bombers, 36 Type 99 dive bombers and 36 Type 0 fighters, 108 aircraft in all, led by Lt. Joichi Tomonaga of Hiryu.

The first contact with the carriers came from a PBY flown by Lt. Howard Ady, about 200 miles northwest of Midway, soon after 0530. NHHC’s photographic history puts it “soon after 0530”; the time usually given in modern accounts that follow Parshall and Tully is 0534. Heinl, writing in 1948 from the records then available, noted that the PBY carrier sighting is variously timed between about 0525 and 0552, depending on the source. Ady’s initial report was brief; his fuller report came later, after he had worked around the force.

A second PBY, flown by Lt. William Chase, saw the strike itself. The ONI narrative times his report at 0545 and gives it as “many planes heading Midway, bearing 320, distance 150.” At 0552, according to the same narrative, a PBY reported two carriers and the main body bearing 320 degrees from Midway at 180 miles, course 135, speed 25. That report was intercepted by Enterprise and passed to Rear Adm. Frank Jack Fletcher and Rear Adm. Raymond A. Spruance, who began working out launch times for their own strikes; their decisions are covered in Chapter 4.

Midway’s radar had the incoming raid at nearly the same moment. Heinl cites the 6th Defense Battalion log as recording planes at 0528 at 100 miles and, at 0555, “many planes, 89 miles, 320°.” The ONI narrative has the air station’s radar picking up the formation at about 0550 at 93 miles and about 10,000 feet. The air-raid alarm sounded at 0555, Condition One was set, and every aircraft that could fly was ordered off the ground. By 0600 or shortly after, according to the ONI narrative, everything flyable was airborne except one J2F utility plane. The fighters went to meet the raid; the bombers and torpedo planes went for the carriers; the remaining PBYs and utility aircraft cleared out of the way.

Getting the whole air garrison off the ground before the bombs fell was the single most important thing Midway did on 4 June. The Japanese raid found empty runways and no aircraft on the ground to destroy, which is why Tomonaga’s crews, for all the damage they did to buildings and fuel, did not accomplish the strike’s principal purpose of eliminating the atoll’s air strength.

VMF-221 against the Zeros

The interception was fought by two groups. Heinl, writing from squadron records, describes Maj. Parks leading a group of eight F2A-3s and five F4F-3s, and Capt. Kirk Armistead following with twelve F2A-3s and one F4F-3 about ten minutes behind. The total that got into the air is one of the minor disagreements of the battle. Heinl reports 26 of the squadron’s 28 aircraft operational and notes that one F2A-3 turned back with engine trouble, so that 25 may be the correct number engaged. The ONI narrative refers to a group of 27 fighters. NHHC’s H-Gram says 21 F2As and seven F4Fs launched and notes that the sources conflict on how many actually got airborne.

Parks’s group made contact at about 0616, some 30 miles out, at around 12,000 feet. According to Heinl, Capt. John F. Carey called the “Tally-ho.” The Marines went for the bombers and at first did not see the escort. The Zeros, above and behind, came down on them almost at once. The ONI narrative records a Midway outpost reporting at 0619 “two planes falling in flames.” The Marine fighters did shoot down some of the bombers; NHHC credits them with three of the Type 97 attack planes. Then the escort took them apart.

Of the twelve pilots in Parks’s group, Heinl records, only three survived: Carey, who was wounded; Capt. Marion E. Carl; and 2d Lt. Clayton M. Canfield. Parks was among those lost. Armistead’s group fared somewhat better but was also badly cut up. When the fight was over, according to the ONI narrative, 15 Marine fighters were missing (13 F2A-3s and two F4F-3s) and seven more were severely damaged. The 1958 Marine Corps history says only ten fighters came back to Midway and only two of those were fit to fly again. NHHC’s H-Gram reaches the same conclusion: 13 F2As and three F4Fs shot down, and only two flyable aircraft among those that recovered.

The fighter pilots’ own claims were very large. The ONI narrative, drawing on their reports, credited the fighters with at least 43 Japanese planes, plus an unknown number shot down by pilots who did not return, and estimated the attackers at more than 100 aircraft including 60 to 80 dive bombers and about 50 fighters. Those claims are not supported by the Japanese record. The squadron’s ordeal does not need them: VMF-221 flew obsolescent fighters into a fight it could not win, in order to give the antiaircraft guns fewer bombers to shoot at, and it paid for that with most of its pilots. The VMF-221 unit page carries the squadron’s roll.

The raid, 0630 to 0700

At 0629, Heinl records, the enemy was eight miles out. At 0630 Col. Shannon gave the order to “open fire when targets are in range,” and at 0631 every antiaircraft battery on Midway opened fire. The first bomb fell at 0630, according to the ONI narrative. The 1958 Marine Corps history describes the first wave of level bombers coming over at about 14,000 feet, with antiaircraft fire knocking down two of them and 22 bombing, and a second wave of 18 dive bombers following immediately after the first bombs fell. Both islands were struck.

On Sand Island, nine 242-kilogram bombs hit the seaplane hangars and set them on fire, and a large fire started in the fuel-oil tanks about 500 yards to the north, according to Heinl and the 1958 history. The ONI narrative logs the hangar hit several times and burning at 0641 and the southwest oil tanks afire at 0644. The fuel storage at the north end of the island and the dispensary were also hit; the ONI narrative records the hospital and the contractors’ canteen burning at 0635 and the laundry hit.

On Eastern Island the bombers went for the airfield’s support facilities. The 1958 history records hits on the Marine mess hall, galley and post exchange; the powerhouse took two heavy bombs and was “virtually destroyed”; and the command post took a direct hit. Maj. William W. Benson, the Eastern Island sector commander, was killed. Four VMF-221 ordnance men were killed by a single direct hit. The Marine history puts total casualties among the ground defenders at 24. The runways themselves were left largely intact, and the ONI narrative notes that damage to the gasoline system meant aircraft had to be refueled by hand for the rest of the battle, an exhausting job that slowed every subsequent sortie.

By 0648, according to the ONI narrative, the attackers were leaving on a bearing of 300 degrees; at 0655 an outpost reported that all enemy planes had left the area, and at 0715 the fighters still airborne were ordered to land and refuel. Almost every above-ground structure on both islands had been destroyed or badly damaged. But the guns were intact, the runways usable, and the aircraft that had been launched at 0555 were either on their way to the Japanese carriers or about to return.

How many Japanese aircraft the raid cost is another figure with several answers. The ONI narrative credited the antiaircraft batteries with ten planes in addition to the fighters’ claims. The 1958 Marine Corps history noted that Japanese authorities admitted only six planes failing to return (three level bombers, two dive bombers and one fighter) while observing that Marine pilots’ reports suggested more. NHHC’s H-Gram 006, reflecting modern research that draws on Parshall and Tully, gives 11 aircraft shot down, 14 severely damaged and 29 others hit, with 20 aviators killed or missing: nearly half the strike lost or damaged in some degree. The differences come partly from counting rules, since aircraft that returned but never flew again are treated differently in different tallies.

One of the men on Sand Island that morning was the Hollywood director John Ford, then a Naval Reserve officer whose photographic unit worked under William J. Donovan’s Office of Strategic Services. In late May, as HistoryNet’s account of the episode relates, Nimitz summoned Ford to Pearl Harbor and gave him what he described as a dangerous assignment: to go to Midway and film the fighting when it came. Ford and his cameraman, Jack Mackenzie Jr., went out by destroyer.

Capt. Simard suggested that Ford station himself on top of the powerhouse, which gave him a view across the island and a telephone line to the command post. During the raid a fragment of concrete struck Ford in the head and briefly knocked him out, and he took a shrapnel wound in the arm, by the HistoryNet account about three inches long. He kept filming. His footage of the bombers over Sand Island, the tracer and the burning fuel tanks, with film shot at sea from Hornet by Lt. Kenneth M. Pier, became the 18-minute Navy documentary The Battle of Midway, released in September 1942. The film’s later history and its awards belong to Chapter 9.

“There is need for a second attack wave”

Tomonaga’s crews had done serious damage, but they had not done what the strike was for. Midway’s aircraft had been in the air when the bombs fell; its guns were still firing as the Japanese withdrew; its runways were usable. Shortly before 0700, as he led the survivors back toward the carriers, Tomonaga radioed Nagumo. The 1958 Marine Corps history renders the message this way:

There is need for a second attack wave.

— Lt. Joichi Tomonaga, strike leader, by radio to Nagumo, about 0700, 4 June 1942 (as given in Hough, Ludwig and Shaw, Pearl Harbor to Guadalcanal, 1958)

NHHC’s H-Gram describes the signal as a code-word message received by Nagumo shortly before 0700. Its significance lies in what Nagumo had kept back. Under the plan, the carriers held a reserve strike force on deck and below, armed with torpedoes and armor-piercing bombs, in case American ships were found. Tomonaga’s report was a request to rearm that reserve for land targets and send it against Midway. It arrived just as Midway’s own aircraft began appearing over the Japanese force, and before any of Nagumo’s scouts had reported American ships. How Nagumo handled the rearming, and the late report from the cruiser Tone‘s No. 4 scout plane, is the subject of Chapter 5.

The island strikes back: Avengers and Marauders

The first of Midway’s attackers to reach the Japanese carriers were the torpedo planes. The six VT-8 Avengers under Fieberling and the four Army B-26s under Capt. James F. Collins Jr. had taken off separately and did not attack as a coordinated group, though they arrived at nearly the same time, about 0705 by the ONI narrative and Craven and Cate. Whether a joint attack had been planned and missed, or whether no real coordination plan existed, is one of the points NHHC’s H-Gram identifies as disputed.

The Avengers went in low against heavy fighter opposition. NHHC records that at least two of them launched torpedoes at Hiryu, which avoided them. Ens. Albert K. “Bert” Earnest, his aircraft badly shot up, launched at the light cruiser Nagara and missed, then nursed the Avenger back to Midway and landed it with one wheel up, as the ONI narrative records. Five of the six TBFs were lost. Of the 18 men who flew the mission, NHHC counts 16 killed; Earnest and one of his crewmen were the only survivors. Fieberling was among the dead.

The B-26s attacked Akagi. The ONI narrative, based on the surviving crews’ reports, describes the aircraft meeting six Zeros head-on, Collins dropping his torpedo from about 800 yards at about 220 feet, and 1st Lt. James P. Muri’s aircraft dropping from about 450 yards at 150 feet. All the torpedoes missed. One B-26 was shot down. Another, badly damaged, nearly struck Akagi‘s bridge before crashing into the sea. NHHC notes that whether that aircraft was out of control or deliberately aimed at the bridge “remains unknown,” and that it may or may not have released its torpedo. Collins and Muri both got back to Midway and crash-landed; Muri’s aircraft, the “Susie-Q,” had been hit more than 500 times, and all three of his gunners were wounded.

The claims are instructive. The ONI narrative records that the B-26 pilots “believed” they had put two torpedoes into a carrier near the bow, and a B-17 pilot reported an Avenger torpedo hit. The Japanese records show no torpedo hits by either group. Craven and Cate, working from Japanese survivors’ testimony, concluded that neither the B-26s nor the TBFs scored. Two of the four B-26s and five of the six TBFs had been lost, and the survivors were wrecked. The Army’s part in the battle is covered further in the Army Air Forces collection, and the Avenger detachment on the VT-8 page.

Henderson’s glide-bombing attack

VMSB-241 flew as two groups. Henderson led the SBD-2s and Maj. Benjamin W. Norris led the SB2U-3s. The ONI narrative records that Henderson’s group was to number 18 aircraft, that two developed engine trouble before takeoff, and that 16 took off between 0610 and 0620 and climbed to 9,000 feet. Norris had 12 Vindicators, one of which turned back with a loose cowling, leaving 11. The National WWII Museum’s account of the squadron gives a similar picture: 30 dive bombers, three kept on the ground by maintenance problems, and 27 sent out in the two groups.

The reason for the glide attack is recorded in the ONI narrative: ten of Henderson’s pilots had joined the squadron only a week earlier, and only three had any experience in the SBD. A true dive-bombing attack, with the aircraft at seventy degrees or steeper, required practice the squadron had not had time to give its newcomers. A glide-bombing approach at a shallower angle was easier to fly and easier to hold in formation, but it kept the aircraft exposed longer to fighters and antiaircraft fire and made the bomb’s aim less certain against a ship turning at speed.

The 1958 Marine Corps history records that Henderson’s group sighted the Japanese ships at 0744 and was immediately attacked by Zeros. The ONI narrative times the attack itself at about 0755. Henderson was hit early and his aircraft went down in flames, and Capt. Elmer G. Glidden took over the lead. The survivors pushed the attack down to 500 feet or less before releasing. Their target, as most modern accounts identify it, was Hiryu, which NHHC describes as straddled by near misses but not hit. The 1943 narrative recorded claims of three direct hits and several near misses on a “Kaga-class” carrier seen smoking badly, with a footnote acknowledging that the identification “possibly was mistaken.”

The cost was half the group. NHHC counts six SBDs shot down and two more so badly damaged that they did not make it back to Midway, eight in all. The ONI narrative records that eight returned and six of those were fit for service; it describes two going down in flames over the target, one going out of control before reaching Midway (its pilot picked up by a PT boat), and one forced down 100 miles west, its pilot and gunner rescued two days later. The rear gunners claimed four Zeros in flames and two probables. NHHC credits the SBD crews with one Zero shot down.

Norris, the Fortresses and the battleship

The last two attacks from Midway that morning arrived after Henderson’s. Sweeney’s B-17s had been in the air since about 0415. The ONI narrative records that they found the enemy at 0732 but did not locate the carriers until 0810, and then attacked from high altitude from the northwest. The number of Fortresses involved is variously given: the ONI narrative says 15 cleared Midway; Craven and Cate say that “never were more than fourteen B-17’s over a group of targets at any one time”; NHHC’s H-Gram gives 12. The ONI loading was eight 500-pound bombs per aircraft, with two aircraft carrying eight 600-pounders.

The ONI narrative records claims of three hits on two carriers, probably two of them on Soryu. None hit. NHHC’s H-Gram says the B-17s dropped more than 320 bombs, a figure that may include more than this one attack; Craven and Cate total 314 bombs for all B-17 sorties between 3 and 5 June. What the photographs taken from the bombers show is the effect the Fortresses did have: carriers swinging through tight circles to throw off the bombardiers. NHHC notes that the Zeros soon gave up attacking the heavily armed bombers, and that the bombing forced the carriers to maneuver in a way that disrupted flight operations.

Norris’s eleven Vindicators came in last, sighting the enemy at about 0820 by the ONI timing. Facing the fighter cover over the carriers and flying slow, vulnerable aircraft, Norris chose to attack the nearest heavy ship, the battleship Haruna, rather than press on in search of a carrier; the 1958 history notes that some of his pilots went after Kirishima nearby. Neither ship was hit. The 1943 narrative recorded claims of two direct hits and two near misses, with the battleship seen smoking and listing. Losses again depend on the source. NHHC counts four: two shot down and two that ditched on the way home. The 1958 Marine Corps history says three were shot down. The ONI narrative mentions two forced down in the water, with two pilots and a gunner rescued.

Taken together, Midway sent between 49 and 52 aircraft against the Japanese carrier force between about 0705 and 0830, depending on the B-17 count, and lost 18 or 19 of them in the attacks themselves, depending on the count of Norris’s losses. The table of claims against results is the starkest single measure of the morning, and it explains why the official accounts of 1942 and 1943 read so differently from the ones written after the war.

What the morning’s attacks achieved

Measured in damage, the answer is nothing. No Midway-based aircraft hit a Japanese carrier on 4 June. Craven and Cate, writing for the Army Air Forces, conceded as much and drew a broader conclusion about their own service’s part: “Midway was not a test of the bomber.” The attacks were small, staggered and uncoordinated; they used three different methods, from low torpedo runs to glide-bombing to high-altitude bombing; and they ran into a fighter patrol that, at that hour, was fresh and well positioned.

Measured in effect, the answer is more complicated, and modern research has made it more so. The attacks forced the Japanese carriers to spend more than an hour maneuvering hard, launching and recovering fighters for the combat air patrol and keeping their decks clear for that purpose. NHHC’s H-Gram notes that the B-17 attack in particular disrupted flight operations. They also arrived immediately after Tomonaga’s report and before Nagumo had any confirmed knowledge of American carriers nearby, reinforcing the case for a second strike against Midway. Parshall and Tully argue that the cumulative effect of these attacks, together with the carrier torpedo attacks that followed, was to keep Nagumo’s decks and his decision-making in a state of constant interruption until the American dive bombers arrived at about 1020. That argument, and the older idea that the Japanese were only “five minutes” from launching when they were hit, are examined in Chapter 5.

The price fell hardest on the Marines. VMF-221 had effectively ceased to exist as a fighting squadron by 0715. VMSB-241 had lost at least eleven aircraft and many of its crews by mid-morning, including its commanding officer. That evening, at about 1900, the 1958 history records, six SBD-2s and five SB2U-3s went out to look for a burning Japanese carrier reported to the northwest. They did not find it. Maj. Norris, who had led the Vindicators that morning, did not return; the National WWII Museum’s account describes his aircraft lost in a spiral dive into the overcast. Fleming helped bring the rest of the group home, guided by oil fires and searchlights on the atoll, according to the National WWII Museum. The squadron’s full record, including the 5 June attack on Mikuma, is on the VMSB-241 unit page.

Midway after the morning

By 0830 Midway’s part in the main action of 4 June was largely over. Its fighters were gone, most of its strike aircraft lost or wrecked, its powerhouse destroyed and its fuel system crippled. With the gasoline system damaged, the surviving aircraft had to be refueled by hand. Yet the atoll had done what Nimitz needed it to do. Its patrol planes had found both the invasion force and the carrier force and reported them in time; its garrison had come through the first strike with its defenses intact; and its counterattacks, however costly, had kept the Japanese carriers busy during the very hour in which the American carriers were launching their own strikes.

Army B-17s continued to operate from the atoll for the rest of the battle. Craven and Cate record further B-17 attacks later on 4 June, including one by six B-17s that found the damaged Hiryu; over three days the Seventh Air Force flew 16 B-17 attacks totaling 55 sorties and one B-26 attack, at a cost of two B-26s with their crews and two B-17s. Hiryu‘s afternoon strikes on Yorktown and her own destruction are covered in Chapter 6. Midway’s last air attacks of the battle, against the cruisers Mikuma and Mogami on 5 June, are covered in Chapter 7.

The invasion never came. Yamamoto cancelled Operation MI early on 5 June, after the loss of all four of Nagumo’s carriers made a landing against a defended atoll impossible. The garrison would spend the following days counting its dead, rescuing downed aircrew from the sea and burying the men killed on 4 June. Those losses, and the men in rafts and boats who were still waiting for rescue days later, are the subject of Chapter 8.

The airfield on Eastern Island was later named Henderson Field in honor of the squadron commander who died leading its first attack, as was the airfield the Marines captured on Guadalcanal two months after Midway. The name was given twice within a year to the officer who led the atoll’s most costly attack of the morning.